Not only naming and wording

To follow up on Daniel,
I absolutely agree that the best way to procede is to promote 
the guidelines as of general or universal acceesibility/usability 
rather than for people with disabilities.

First, because it is corrrect. Second, because it sells better.

One type of agents that could be mentioned more often in
many particular guidelines are indexing robots.

In my view it requires a little bit of content changing as well.
If it will be only in the name and wording, readers will find 
quickly that they have been cheated.

I would be careful in suggesting techniques that are usability 
increasing in some particular situations but are universally 
usability reducing. 

For example the technique of not nesting unordered lists and 
using ordered lists insted, even if the content calls for unordered 
ones. This goes against the spirit of universal design as it violates 
the principle of using the right structural element for the content 
in question.

Another problematic technique is the so called invisible D-links,
which are as invisible as the "HTML terorists' " invisible images.
I do not think that this hack should be part of a W3C recommendation.

If a description is vital for everybody who doesn't see images, 
it should not be invisible, and should have a meaningful link 
description, rather than the letter D (which is meaningfull 
only to accessibility activists in English speaking countries, 
but not that much to normal people who turn off image loading).

If it is only suplementary information that is inessential 
for understanding the page a longdesc should do, even 
though few browsers support this now.

Most users in the world get highly confused when they encouter 
a site full of Ds lying all over the place; and end up following links
that they can't figure out what they are for even after following four 
or five of them; this technique is a major usability hazard that no serious 
web designer interested in a wide readership would implement. Since one 
can't realy make them invisible, they shouldn't be there at all.

With longdesc it is a different story as (good) browser's default 
should be hide longdesc, and the knowledgable user who 
wants them will configure the browser accordingly.

Nir Dagan, Ph.D.
http://www.nirdagan.com
mailto:nir@nirdagan.com

"There is nothing quite so practical as a good theory." 
-- A. Einstein

Received on Monday, 23 November 1998 12:02:17 UTC